legacycentral bbs

I'm wondering if I have a failing pressure sensor...
Some times when I get into boost, about 1psi, I get a hard fuel/boost cut. It is pretty irregular although I'm starting to think it has to do with heat because it comes after driving for while if it is cool out, like int the morning, and it comes almost immediatly if the car has sat in the sun all day.

So I'm wondering if I have a sensor that is bad or if I have a connection that is going bad so that the resistance is raising with heat?

I would appreciate any ideas that any one has.

I thought maybe it was my water temp sensor at first because it has a rough idle some times but at other times it idles smoothly. It usually idles roughly if I start the car up after I have already been driving it for some time.

Does any one know if the pressure sensor can cause the car to idle roughly?

Thanks for any help,
Noah

Do you get any CEL codes?

The pressure sensor shouldn't affect idle.

Do you have access to a Select Monitor? If not, you should build my scantool:

http://www.graphics.cornell.edu/~v/bcbfscan/

Then you can watch the pressure sensor readings, as well as the water temperature readings.

I don't have the car wired to use a scan tool at this point but it sure would be helpful right now.

Good to know the pressure sensor should not affect idle.

the rough idle could also be due to the fact that we adjusted the set screw for the throttle plate and the idle air control valve settings.

Dave made one of your scan tools and was showing it to me. It was very cool.

Do you know when, or under what conditions, the pressure exchange solenoid switchs between out side pressure and manifold pressure? Does it do it during boost and vacume or just during vacume?

You actually only need to connect two wires to the ECU. My ECU pinouts page tells you which ones (the Select Monitor Tx and Rx pins):

http://www.graphics.cornell.edu/~v/ecupins/

I think you should hook those wires up. You have the opportunity to get a lot of help diagnosing engine management problems.

Playing with the throttle plate and IAC valve calibrations certainly seems like it could affect idle quality.

When the pressure exchange solenoid is off, the sensor measures manifold pressure. When it's on, it measures atmospheric pressure.

The ECU normally leaves the solenoid off. When it does decide to sample atmospheric pressure, it activates the solenoid for about half a second.

The ECU samples atmospheric pressure every couple of minutes (roughly), and also every time you transition from being on vacuum to being on boost.

Do you get a Check Engine light?

I do have a couple of check engine lights and may have a few more now but I have not checked what they are in a while but the codes I had were 11, 24, and 35. I only remember that one of them had to do with an emissions solenoid.

I'm going to build the scan tool tomorrow and try and check the check engine codes again.

thanks for the help.

Hmm, those are respectively codes for the crank angle sensor, IAC valve, and purge valve... none for the pressure sensing stuff.

Phil (BAC5.2) says that fuel cut is sometimes a symptom of the IAC valve code though. I haven't observed that myself, but maybe that's what you're seeing.

Did you clear the codes to see which ones returned?

I checked and tried to clear the codes when I first got the motor running. I got the codes but had no luck clearing them. Is there a way to clear the codes with the ECU out of the car?

I may try a different IAC today since that is something I have easy access to.

I will also try and get the scan tool built so I can watch everything and maybe see what is happening when the fuel cut happens. If it happens after I try another IAC that is.

Thanks for the help!!!

Noah

Just having the ECU disconnected from its harness for a while ought to clear its memory.

When you tried clearing them by putting the ECU in clear memory mode, did it give you the "all clear" code, or did it give you the trouble codes?

I just got the trouble codes again. If I remember correctly to clear the codes we jumpered a couple of wires and then followed the directions in the book which had us driving around at a certain speed or rpm... It has been almost six months though so I don't remember exactly what we did. But it didn't work.

Yeah, that sounds right...

When you try to clear codes, the ECU actually does an active check of the systems before clearing them. If it detects a fault it won't go through with the code-clearing process.

A Select Monitor can force the codes to be cleared anyway, but they may just come right back (assuming the ECU itself is functioning correctly, they will come back). My scantool doesn't have that capability, though.

swapped out the IAC and no luck with the fuel cut...
so next is getting the scan tool built and watching to see what happens...

When the IAC is going, Fuel cut is the saftey net and is based on the ECU's adaptive features. When any of a few variables dramatically change beyond a set of limits (only above).

Drive hard from the time your car is warmed up, for the next few miles. Then, unless you drive like a grandma for a while, you SHOULDN'T hit fuel cut.

It seems to me that the ECU is adaptave (which it is) and "resets" it's memory of driving style every time you shut off the car (which is why one day, the car seems really fast, and the next it seems like it runs like shit). This would explain my erratic Fuel Cut (like the other day, I hit fuel cut whenever I tried to pull in 4th. Today, I was going WOT in 3rd and 4th, and didn't hit a single hiccup.)

Vikash will correct that last part if it's wrong.

I don't know whether it's right or wrong. It doesn't seem sensible that it would forget every time you shut the engine off, but I don't know...

Also, like I said, I've never experienced IAC-related fuel cut.

I hit fuel cut when first entering R-Mode and hitting boost right away, it's a VERY violent cut too, not like IAC related cut, that feels like a soft rev limiter.

If it doesn't forget, then I don't know how to explain my erratic car behavior, other than a bad ECU. Car drives AWESOME one day, and horrible the next.

My fuel cut is very strong, not soft at all.
Mine goes something like this: Morning car drives well, boost comes on consistantly and is strong until I have to stop a few times and then as soon as I hit about 1psi strong fuel cut.

I thought maybe it was the water temp sensor because it seems like a heat related issue but from the testing I did the temp sensor seems to be working properly.

I still think it is tempature related. I'm thinking maybe one of connections is just bad enough that as things get hotter the resistance goes up and the ECU no longer gets the right signal. That is why I'm hoping to get the scan tool built and running so I can watch what is happening with some of the inputs.

If I'm correct and I could be wrong. your ECU should have a constant power so that it does not forget the learning that it does unless that power is taken away for a period of time. That is the reason you can restet your ECU by disconnecting it for a while.

I did find a bit of info on the IAC code #24: "Air control valve. Air control valve inoperative (Abnormal signal produced in monitor circuit) Prevents abnormal engine speed using "fuel cut" in relation to engine speed, vehicle speed, and throttle sensor position."
I don't remember where I got that quote but it is not from the Subaru manual (No. 2 & 3) I have. All it says for "Trouble Code (24) is -Erroneous idiling, Engine stall, and Engine breathing.

If you go by the unknown source it seems like IAC could be a cause for the issue we are having. Although you said you get a soft cut and all I get is a hard cut. So we must have some different issues with our cars.

If you want to test the IAC check resistance (on IAC itself) between termanils 1-2 and 2-3 and the resistance should be 9 ohms.

Time for bed

I found the problem with boost that I was having was because the pressure sensor was bad. I found out by borrowing a known good one.

I also got a cable built and the scan tool software working. Thanks for cool software! Intially I could not get it to work until I made a bootable cd and then it worked perfectly.

I have some results from idle and I'm looking forward to recording results while driving, when I have some help of course.

Rom ID 72.36.22 (91-94 turbo)

15 deg BTDC @ idle
Air flow .82-.88
Load 30
TPS 4.72V
Injector pulse width 2.048-2.176
IAC 34 %
Fuel trim 7.81%
Timing cor 0%
Boost con duty cyc 1.56%

Do these numbers look normal?

Hmm, yeah, I thought it had to do with the pressure sensing stuff.

If you don't want to find and pay for an OEM sensor, you can adapt other sensors. I've done it with a cheap and common Chrysler sensor:

http://www.graphics.cornell.edu/~v/chrysler_map/

I'm glad the scan tool worked for you... if you tried running the .COM file under any kind of operating system more sophisticated than DOS it definitely wouldn't work right. But that's what the bootable disk/CD form is for.

The numbers all look okay except that the fuel trim seems a little high. Not a lot high, but a little high. But you're running an EJ20G with an EJ22T ECU, right? So maybe that's reasonable.

vrg3 wrote: The numbers all look okay except that the fuel trim seems a little high. Not a lot high, but a little high. But you're running an EJ20G with an EJ22T ECU, right? So maybe that's reasonable.

What is normally seen with the turbo Legacy for fuel trim?

I actually have an even more convoluted set up then you mention above. I'm using a 2.5 short block with the EJ20G heads and 440cc injectors. So that more than likely has something to do with the change in fuel trim.

Is there a web site or post that has values for the car under certain conditions? For example I would like to know what would normally be seen for fuel trim on the 2.2T under different conditions.

I would also be interested to know what the highest Ign. Timing people have seen is? I saw just over 50 once.

Thanks again.

At idle you usually get plus or minus a couple percent.

Using those injectors definitely means your fuel maps will be wrong, so there's no mystery. I'm curious -- why aren't you using correctly-sized injectors for the ECU you're using?

A couple of people have started threads here about their scantool readings, so you could try searching for them.

The injectors came with the EJ20G and it is the only set I have at this point. I have been considering trying to find some stock size injectors but I don't really want to spend money on injectors that I won't use for long. I plan on getting another turbo and tuning the car soon.

Ah, gotcha...

Maybe you could get an adjustable FPR to compensate temporarily, if such a device would also be useful when you begin tuning.